Setup and data wrangling

Load necessary packages

Show code

library(gtsummary)
library(survival)
library(readxl)
library(dplyr)
library(lubridate)
library(readr)
library(survminer)
library(ggsurvfit)
library(rms)
library(gt)
library(bstfun)
library(swimplot)
library(ggplot2)
library(RColorBrewer)

Read in the data

Show code

db1 = read_excel(path = "9762 data for analysis 1.23.24.xlsx", sheet = "Survival")

dataset <- db1 %>%
  mutate(
    dx_dt = ymd(dx_dt),
    infuse_dt = ymd(infuse_dt),
    birthdate = ymd(birthdate),
    dt_relapse = ymd(dt_relapse),
    Death.or.DLC = ymd(Death.or.DLC),
    Death.or.relapse.or.DLC = ymd(Death.or.relapse.or.DLC),
    Days.to.death.or.DLC = Death.or.DLC - infuse_dt,
    Days.to.death.or.relapse.or.DLC = Death.or.relapse.or.DLC - infuse_dt,
    Days.from.dx.to.infusion = infuse_dt - dx_dt,
    Age = (infuse_dt-birthdate)/365
  )

Data tables

Table 1 reproduction

Show code

theme_gtsummary_compact()


dataset %>%
  transmute(
    "Age (years)" = Age,
    "Female sex" = ifelse(sex == "f", 1, 0),
    "Race" = recode(race, "asian" = "Asian", "white" = "White"),
    "ECOG performance status" = ecog,
    "Median time from diagnosis to CAR T cell therapy" = Days.from.dx.to.infusion/365,
    "Disease subtype" = factor(subtype_category, levels = c("IgA", "IgG", "IgM", "Light chain", "Oligosecretory", "Primary plasma cell leukemia")),
    "ISS disease stage at diagnosis" = iss,
    "Revised ISS disease stage at diagnosis" = riss,
    "ISS disease stage at enrollment" = iss_enrollment,
    "Revised ISS disease stage at enrollment" = riss_enrollment,
    #"High-risk cytogenetic abnormality at any time" = HRCA,
    "t(4;14) any" = t414_any,
    "t(14;16) any" = t1416_any,
    "del(17p) any" = del17p_any,
    "amp(1q) any" = amp1q_any,
    "High bone marrow tumor burden" = ifelse(High.tumor.burden == 1, ">30%", "10-30%"),
    "Extramedullary disease" = extramedullary,
    "Circulating plasma cells at screening" = cpcs,
    "Tumor BCMA expression (ABC)" = bcma_scrn,
    "Number of previous antimyeoma regimens" = factor(case_when(
      no_priortx <=6 ~ "4-6",
      no_priortx >6 & no_priortx <=9 ~ "7-9",
      no_priortx >9 ~ "10 or more"), levels = c("4-6", "7-9", "10 or more") ),
    "Previous autologous HCT" = factor(case_when(
      no_auto == 0 ~ "0",
      no_auto == 1 ~ "1",
      no_auto >1 ~ ">1"), levels = c("0","1",">1") ),
    "Previous allogeneic HCT" = allo
  ) %>%
  
  tbl_summary(
    missing = "ifany",
    statistic = list(
      all_continuous() ~ "{median} ({min} to {max})",
      all_categorical() ~ "{n} ({p}%)"
    )
  ) %>%
  bold_labels() %>%
  add_variable_grouping(
    "High-risk cytogenetic abnormality at any time" = c("t(4;14) any", "t(14;16) any", "del(17p) any", "amp(1q) any")
  )

Characteristic	N = 31¹
Age (years)	64 (38 to 77)
Unknown	6
Female sex	9 (36%)
Unknown	6
Race
Asian	3 (12%)
White	22 (88%)
Unknown	6
ECOG performance status
0	5 (20%)
1	17 (68%)
2	3 (12%)
Unknown	6
Median time from diagnosis to CAR T cell therapy	5.1 (1.6 to 14.6)
Unknown	6
Disease subtype
IgA	4 (16%)
IgG	13 (52%)
IgM	2 (8.0%)
Light chain	3 (12%)
Oligosecretory	1 (4.0%)
Primary plasma cell leukemia	2 (8.0%)
Unknown	6
ISS disease stage at diagnosis
1	4 (20%)
2	6 (30%)
3	10 (50%)
Unknown	11
Revised ISS disease stage at diagnosis
1	2 (18%)
2	3 (27%)
3	6 (55%)
Unknown	20
ISS disease stage at enrollment
1	12 (63%)
2	5 (26%)
3	2 (11%)
Unknown	12
Revised ISS disease stage at enrollment
1	6 (32%)
2	11 (58%)
3	2 (11%)
Unknown	12
High-risk cytogenetic abnormality at any time
t(4;14) any	2 (8.0%)
Unknown	6
t(14;16) any	2 (8.0%)
Unknown	6
del(17p) any	13 (52%)
Unknown	6
amp(1q) any	3 (12%)
Unknown	6
High bone marrow tumor burden
>30%	17 (68%)
10-30%	8 (32%)
Unknown	6
Extramedullary disease	11 (44%)
Unknown	6
Circulating plasma cells at screening	5 (20%)
Unknown	6
Tumor BCMA expression (ABC)	620 (175 to 5,521)
Unknown	6
Number of previous antimyeoma regimens
4-6	6 (24%)
7-9	13 (52%)
10 or more	6 (24%)
Unknown	6
Previous autologous HCT
0	5 (20%)
1	13 (52%)
>1	7 (28%)
Unknown	6
Previous allogeneic HCT	5 (20%)
Unknown	6
¹ Median (Range); n (%)

Swimmer’s plot

Figure 1: Swimmer’s plot

Show code

swimmer_data = as.data.frame(read_excel(path = "9762 swimmer.xlsx", sheet = "Response")) %>%
  mutate(
    Response_start = Response_start/30,
    Response_end = Response_end/30,
    Arrow_start = Arrow_start/30,
    Time = Time/30,
    Response = factor(Response, levels=c("PD","sCR","CR","VGPR","PR","SD"))
  )

pastel1_palette <- brewer.pal(8, "Pastel1")[1:6]


swimmer_plot(df=filter(swimmer_data, !is.na(Response) ), 
            id="Number", 
            end= "Response_end", 
            name_fill="Response",
            stratify = "Dose_level",
            ) +
  scale_fill_manual(values = pastel1_palette) +
  swimmer_arrows(df_arrows = filter(swimmer_data,!is.na(Arrow_start) ),id="Number",arrow_start="Arrow_start", name_col = 'Response', type = "open", cex=1, arrow_positions = c(1,5)) + 
  swimmer_points(df_points= filter(swimmer_data,!is.na(Event) ) ,id='Number',name_shape = "Event", time='Time',size=2,fill='white',col="black") +
  scale_y_continuous(name = "Time since infusion (Months)",breaks = seq(0,10000,by=12)) +
  guides(fill = guide_legend(title = "Response"), colour = FALSE)  # Hide the legend for arrows

Supplemental figure

sBCMA plots

Show code

sBCMA_data = as.data.frame(read_excel(path = "sBCMA data.xlsx"))


ggplot(sBCMA_data, aes(x = timepoint/30, y = sBCMA, color = Tumor.burden)) +
  geom_point(shape = 16, size = 2) +
  theme_classic() +
  labs(x = "Time (months)", y = "sBCMA") +
  scale_color_discrete(name = "Tumor burden") +
  facet_wrap(~Number, scales = "free_y", ncol = 5) +  # Facet by the "Number" variable with free y-axes and 2 columns
  scale_y_log10(limits = c(0.01, 10000)) +
  scale_x_continuous(breaks = seq(0, 48, by = 12)) +
  theme(strip.text = element_text(size = 9)) + 
  geom_vline(aes(xintercept = Relapse.time/30), linetype = "dashed", color = "black")  # Add vertical line

Survival analysis

Length of follow-up

Show code

data1 = dataset %>%
  mutate(
    Reverse_death = ifelse(death == 1, 0,1)
  )

#quantile(prodlim(Hist(time = Real.days.DLC/30, Real.death )~1, data = data1, reverse = TRUE ) )


reverse_km_OS <- survfit(Surv(Days.to.death.or.DLC/30, Reverse_death) ~ 1, data1)
reverse_km_OS

Call: survfit(formula = Surv(Days.to.death.or.DLC/30, Reverse_death) ~ 
    1, data = data1)

   6 observations deleted due to missingness 
      n events median 0.95LCL 0.95UCL
[1,] 25      6   58.6    47.4      NA

Create KM for OS and PFS

Show code

km_OS <- survfit(Surv(Days.to.death.or.DLC/30, death) ~ 1, data = dataset)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
) 

OS$plot <- OS$plot + geom_hline(yintercept = 0.5, linetype = "dashed", color = "grey") +  annotate("text", x = med_OS$median + 0.5, y = 0.54, label = paste(signif(med_OS$median, 2), "(95% CI, ", signif(med_OS$lower, 2), "to", signif(med_OS$upper, 2),")"), hjust = 0,  size = 5)
OS

Show code

km_OS

Call: survfit(formula = Surv(Days.to.death.or.DLC/30, death) ~ 1, data = dataset)

   6 observations deleted due to missingness 
      n events median 0.95LCL 0.95UCL
[1,] 25     19   32.1    20.4      NA

Show code

summary(km_OS,times=c(12))

Call: survfit(formula = Surv(Days.to.death.or.DLC/30, death) ~ 1, data = dataset)

6 observations deleted due to missingness 
 time n.risk n.event survival std.err lower 95% CI upper 95% CI
   12     22       3     0.88   0.065        0.761            1

Show code

km_PFS <- survfit(Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ 1, data = dataset)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS$plot <- PFS$plot + geom_hline(yintercept = 0.5, linetype = "dashed", color = "grey") +  annotate("text", x = med_PFS$median + 0.5, y = 0.54, label = paste(signif(med_PFS$median, 2), "(95% CI, ", signif(med_PFS$lower, 2), "to", signif(med_PFS$upper, 2),")"), hjust = 0,  size = 5)
PFS

Show code

km_PFS

Call: survfit(formula = Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ 
    1, data = dataset)

   6 observations deleted due to missingness 
      n events median 0.95LCL 0.95UCL
[1,] 25     23     18    8.73    27.5

Show code

summary(km_PFS,times=c(12))

Call: survfit(formula = Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ 
    1, data = dataset)

6 observations deleted due to missingness 
 time n.risk n.event survival std.err lower 95% CI upper 95% CI
   12     16       9     0.64   0.096        0.477        0.859

PFS and OS on the same plot

Show code

OS_EFS <- ggsurvplot_combine(list(km_PFS, km_OS),
           data=dataset,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.3,
           tables.theme = theme_cleantable(),
           surv.median.line = "hv", # Specify median survival
           ggtheme = theme_classic()+
            theme(
              axis.text.x = element_text(size = 12),  # Increase x-axis label font size
              axis.text.y = element_text(size = 12),  # Increase y-axis label font size
              axis.title.x = element_text(size = 14),  # Increase x-axis title font size
              axis.title.y = element_text(size = 14),  # Increase y-axis title font size
              plot.title = element_text(size = 12),    # Increase plot title font size
              legend.text = element_text(size = 12),    # Increase legend text font size
              legend.title = element_text(size = 12)    # Increase legend title font size
            ),
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "PFS & OS (%)",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           #fun=formula,
           legend.labs = c("PFS","OS"),
           surv.scale = "percent",
           censor=FALSE
           )

OS_EFS$plot <- OS_EFS$plot + 
  geom_hline(yintercept = 0.5, linetype = "dashed", color = "grey") + 
  geom_vline(xintercept = median(km_OS)[1], linetype = "dashed", color = "grey") + 
  geom_vline(xintercept = median(km_PFS)[1], linetype = "dashed", color = "grey") + 
  annotate("text", x = med_OS$median + 0.5, y = 0.85, label = paste(signif(med_OS$median, 2), "(", signif(med_OS$lower, 2), "-", signif(med_OS$upper, 2),")"), hjust = 0,  size = 5) +
  annotate("text", x = med_PFS$median + 0.5, y = 0.95, label = paste(signif(med_PFS$median, 2), "(", signif(med_PFS$lower, 2), "-", signif(med_PFS$upper, 2),")"), hjust = 0,  size = 5)

OS_EFS

Create KM for DOR

Show code

km_DOR <- survfit(Surv((Days.to.death.or.relapse.or.DLC-28)/30, relapse) ~ 1, data = dataset)
med_DOR <- surv_median(km_DOR)

DOR <- ggsurvplot(km_DOR,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Duration of response",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

DOR$plot <- DOR$plot + geom_hline(yintercept = 0.5, linetype = "dashed", color = "grey") +  annotate("text", x = med_DOR$median + 0.5, y = 0.54, label = paste(signif(med_DOR$median, 2), "(95% CI, ", signif(med_DOR$lower, 2), "to", signif(med_DOR$upper, 2),")"), hjust = 0,  size = 5)
DOR

Show code

km_DOR

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC - 28)/30, 
    relapse) ~ 1, data = dataset)

   6 observations deleted due to missingness 
      n events median 0.95LCL 0.95UCL
[1,] 25     23   17.1     7.8    26.6

Show code

summary(km_DOR,times=c(12))

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC - 28)/30, 
    relapse) ~ 1, data = dataset)

6 observations deleted due to missingness 
 time n.risk n.event survival std.err lower 95% CI upper 95% CI
   12     14      11     0.56  0.0993        0.396        0.793

Stratified survival analysis

Create KM for OS and PFS, stratify by HRCA

Show code

km_OS <- survfit(Surv(Days.to.death.or.DLC/30, death) ~ HRCA, data = dataset)
km_OS

Call: survfit(formula = Surv(Days.to.death.or.DLC/30, death) ~ HRCA, 
    data = dataset)

   6 observations deleted due to missingness 
        n events median 0.95LCL 0.95UCL
HRCA=0 10      7   32.1    18.5      NA
HRCA=1 15     12   27.5    17.7      NA

Show code

med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black"),
           legend.labs = c("Standard","High risk")
) 

#OS$plot <- OS$plot + geom_hline(yintercept = 0.5, linetype = "dashed", color = "grey") +  annotate("text", x = med_OS$median + 0.5, y = 0.54, label = paste("Median =", signif(med_OS$median, 2), "(95% CI, ", signif(med_OS$lower, 2), "to", signif(med_OS$upper, 2),")"), hjust = 0,  size = 5)

OS

Show code

km_PFS <- survfit(Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ HRCA, data = dataset)
km_PFS

Call: survfit(formula = Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ 
    HRCA, data = dataset)

   6 observations deleted due to missingness 
        n events median 0.95LCL 0.95UCL
HRCA=0 10      9   20.6    7.33      NA
HRCA=1 15     14   14.0    8.73      NA

Show code

med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black"),
           legend.labs = c("Standard","High risk")
)

#PFS$plot <- PFS$plot + geom_hline(yintercept = 0.5, linetype = "dashed", color = "grey") +  annotate("text", x = med_PFS$median + 0.5, y = 0.54, label = paste("Median =", signif(med_PFS$median, 2), "(95% CI, ", signif(med_PFS$lower, 2), "to", signif(med_PFS$upper, 2),")"), hjust = 0,  size = 5)

PFS

KM of OS & PFS, stratified by cell dose

Show code

km_PFS <- survfit(Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ intended_cell_dose, data = dataset)
km_PFS

Call: survfit(formula = Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ 
    intended_cell_dose, data = dataset)

   6 observations deleted due to missingness 
                                        n events median 0.95LCL
intended_cell_dose=Dose level 1 - 50E6  7      6  18.00    5.87
intended_cell_dose=Dose level 2 - 150E6 8      7  24.32   12.20
intended_cell_dose=Dose level 3 - 300E6 7      7  23.20   12.63
intended_cell_dose=Dose level 4 - 450E6 3      3   7.87    3.07
                                        0.95UCL
intended_cell_dose=Dose level 1 - 50E6       NA
intended_cell_dose=Dose level 2 - 150E6      NA
intended_cell_dose=Dose level 3 - 300E6      NA
intended_cell_dose=Dose level 4 - 450E6      NA

Show code

med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.4,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black"),
           #legend.labs = c("Standard","High risk")
)

PFS

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ intended_cell_dose, data = data1)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.4,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ intended_cell_dose, 
    data = data1)

   6 observations deleted due to missingness 
                                        n events median 0.95LCL
intended_cell_dose=Dose level 1 - 50E6  7      7   25.4    10.1
intended_cell_dose=Dose level 2 - 150E6 8      4   58.4    27.5
intended_cell_dose=Dose level 3 - 300E6 7      5   32.1    20.4
intended_cell_dose=Dose level 4 - 450E6 3      3   18.5    16.6
                                        0.95UCL
intended_cell_dose=Dose level 1 - 50E6       NA
intended_cell_dose=Dose level 2 - 150E6      NA
intended_cell_dose=Dose level 3 - 300E6      NA
intended_cell_dose=Dose level 4 - 450E6      NA

KM of PFS, dose level 1-3 vs 4

Show code

data1 <- dataset %>%
  mutate(
    dose_level = ifelse(intended_cell_dose == "Dose level 1 - 50E6","1","2-4")
  )

km_PFS <- survfit(Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ dose_level, data = data1)
km_PFS

Call: survfit(formula = Surv(Days.to.death.or.relapse.or.DLC/30, relapse) ~ 
    dose_level, data = data1)

   6 observations deleted due to missingness 
                n events median 0.95LCL 0.95UCL
dose_level=1    7      6   18.0    5.87      NA
dose_level=2-4 18     17   17.5    8.73    31.7

Show code

med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,60), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Level 1-3","Level 4"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black"),
           pval.coord = c(47, 0.8)
)
PFS

Create KM for OS, stratified by sCR/CR vs VGPR/PR

Show code

data1 <- dataset %>%
  mutate(
    response.category = recode(best_response, "scr" = "sCR/CR", "cr" = "sCR/CR", "vgpr" = "VGPR/PR", "pr"="VGPR/PR")
  )

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ response.category, data = data1)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("sCR/CR","VGPR/PR"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ response.category, 
    data = data1)

   6 observations deleted due to missingness 
                           n events median 0.95LCL 0.95UCL
response.category=sCR/CR  17     12   35.5    27.5      NA
response.category=VGPR/PR  8      7   19.1    17.1      NA

Show code

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC-28)/30, relapse) ~ response.category, data = data1)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("sCR/CR","VGPR/PR"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC - 28)/30, 
    relapse) ~ response.category, data = data1)

   6 observations deleted due to missingness 
                           n events median 0.95LCL 0.95UCL
response.category=sCR/CR  17     15  22.27   17.07    30.8
response.category=VGPR/PR  8      8   7.28    6.93      NA

Create KM for OS & PFS, stratified by treatment lines <8 vs >=8

Show code

data1 <- dataset %>%
  mutate(
    prior.lines.8 = ifelse(no_priortx >=8, 1,0)
  )

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC)/30, relapse) ~ prior.lines.8, data = data1)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("<8","\u22658"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ prior.lines.8, data = data1)

   6 observations deleted due to missingness 
                 n events median 0.95LCL 0.95UCL
prior.lines.8=0 11     10   12.6    8.37      NA
prior.lines.8=1 14     13   21.1   12.20      NA

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ prior.lines.8, data = data1)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("<8","\u22658"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ prior.lines.8, 
    data = data1)

   6 observations deleted due to missingness 
                 n events median 0.95LCL 0.95UCL
prior.lines.8=0 11      7   20.4    17.7      NA
prior.lines.8=1 14     12   33.0    25.4      NA

Create KM for OS & PFS, stratified by hx of prior allo

Show code

data1 <- dataset %>%
  mutate(
    prior.lines.8 = ifelse(no_priortx >=8, 1,0)
  )

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC)/30, relapse) ~ allo, data = data1)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ allo, data = data1)

   6 observations deleted due to missingness 
        n events median 0.95LCL 0.95UCL
allo=0 20     19   16.0    8.07    28.2
allo=1  5      4   21.1   12.20      NA

Show code

summary(km_PFS,times=c(12))

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ allo, data = data1)

6 observations deleted due to missingness 
                allo=0 
        time       n.risk      n.event     survival      std.err 
      12.000       12.000        8.000        0.600        0.110 
lower 95% CI upper 95% CI 
       0.420        0.858 

                allo=1 
        time       n.risk      n.event     survival      std.err 
      12.000        4.000        1.000        0.800        0.179 
lower 95% CI upper 95% CI 
       0.516        1.000

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ allo, data = data1)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ allo, 
    data = data1)

   6 observations deleted due to missingness 
        n events median 0.95LCL 0.95UCL
allo=0 20     15   27.5    20.4      NA
allo=1  5      4   61.5    17.7      NA

Show code

summary(km_OS,times=c(12))

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ allo, 
    data = data1)

6 observations deleted due to missingness 
                allo=0 
        time       n.risk      n.event     survival      std.err 
     12.0000      17.0000       3.0000       0.8500       0.0798 
lower 95% CI upper 95% CI 
      0.7071       1.0000 

                allo=1 
        time       n.risk      n.event     survival      std.err 
          12            5            0            1            0 
lower 95% CI upper 95% CI 
           1            1

Create KM for OS & PFS, stratified by receipt of prior BCMA-directed therapy

Show code

data1 <- dataset %>%
  mutate(
    prior.lines.8 = ifelse(no_priortx >=8, 1,0)
  )

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC)/30, relapse) ~ any_bcma_tx, data = data1)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ any_bcma_tx, data = data1)

   6 observations deleted due to missingness 
               n events median 0.95LCL 0.95UCL
any_bcma_tx=0 22     20   18.0    8.73    28.2
any_bcma_tx=1  3      3   12.2    5.87      NA

Show code

summary(km_PFS,times=c(12))

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ any_bcma_tx, data = data1)

6 observations deleted due to missingness 
                any_bcma_tx=0 
        time       n.risk      n.event     survival      std.err 
      12.000       14.000        8.000        0.636        0.103 
lower 95% CI upper 95% CI 
       0.464        0.873 

                any_bcma_tx=1 
        time       n.risk      n.event     survival      std.err 
      12.000        2.000        1.000        0.667        0.272 
lower 95% CI upper 95% CI 
       0.300        1.000

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ any_bcma_tx, data = data1)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ any_bcma_tx, 
    data = data1)

   6 observations deleted due to missingness 
               n events median 0.95LCL 0.95UCL
any_bcma_tx=0 22     17   27.5    20.4      NA
any_bcma_tx=1  3      2   32.1    10.1      NA

Show code

summary(km_OS,times=c(12))

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ any_bcma_tx, 
    data = data1)

6 observations deleted due to missingness 
                any_bcma_tx=0 
        time       n.risk      n.event     survival      std.err 
     12.0000      20.0000       2.0000       0.9091       0.0613 
lower 95% CI upper 95% CI 
      0.7966       1.0000 

                any_bcma_tx=1 
        time       n.risk      n.event     survival      std.err 
      12.000        2.000        1.000        0.667        0.272 
lower 95% CI upper 95% CI 
       0.300        1.000

Create KM for OS & PFS, stratified by triple refractory disease

Show code

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC)/30, relapse) ~ triple_ref, data = dataset)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ triple_ref, data = dataset)

   6 observations deleted due to missingness 
              n events median 0.95LCL 0.95UCL
triple_ref=0  5      5   12.2    8.37      NA
triple_ref=1 20     18   18.0    8.73    28.2

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ triple_ref, data = dataset)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ triple_ref, 
    data = dataset)

   6 observations deleted due to missingness 
              n events median 0.95LCL 0.95UCL
triple_ref=0  5      4   27.5    18.5      NA
triple_ref=1 20     15   32.1    20.4      NA

Create KM for OS & PFS, stratified by penta refractory disease

Show code

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC)/30, relapse) ~ pentaref, data = dataset)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ pentaref, data = dataset)

   6 observations deleted due to missingness 
            n events median 0.95LCL 0.95UCL
pentaref=0  8      8   16.0   12.20      NA
pentaref=1 17     15   21.1    8.07    37.4

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ triple_ref, data = dataset)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ triple_ref, 
    data = dataset)

   6 observations deleted due to missingness 
              n events median 0.95LCL 0.95UCL
triple_ref=0  5      4   27.5    18.5      NA
triple_ref=1 20     15   32.1    20.4      NA

Create KM for OS & PFS, stratified by tumor burden (BMPCs <=30 vs >30)

Show code

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC)/30, relapse) ~ High.tumor.burden, data = dataset)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Low","High"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ High.tumor.burden, data = dataset)

   6 observations deleted due to missingness 
                     n events median 0.95LCL 0.95UCL
High.tumor.burden=0  8      7   24.3   18.00      NA
High.tumor.burden=1 17     16   12.6    8.37    27.1

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ High.tumor.burden, data = dataset)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Low","High"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ High.tumor.burden, 
    data = dataset)

   6 observations deleted due to missingness 
                     n events median 0.95LCL 0.95UCL
High.tumor.burden=0  8      5   48.6    33.0      NA
High.tumor.burden=1 17     14   22.8    17.7      NA

Create KM for OS & PFS, stratified by sBCMA at day 90 (< vs >= median)

Show code

data1 <- dataset %>%
  filter(!is.na(sBCMA.d90))%>%
  filter(Days.to.death.or.relapse.or.DLC>=90) %>%
  mutate(
    sBCMA.high = ifelse(sBCMA.d90 >= median(sBCMA.d90),1,0)
  )

median(data1$sBCMA.d90)

[1] 2.915

Show code

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC-90)/30, relapse) ~ sBCMA.high, data = data1)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Low","High"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC - 90)/30, 
    relapse) ~ sBCMA.high, data = data1)

              n events median 0.95LCL 0.95UCL
sBCMA.high=0 10     10  21.80   15.00      NA
sBCMA.high=1 10      8   7.28    4.33      NA

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC-90)/30, death) ~ sBCMA.high, data = data1)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Low","High"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC - 90)/30, death) ~ 
    sBCMA.high, data = data1)

              n events median 0.95LCL 0.95UCL
sBCMA.high=0 10      7   55.4    30.0      NA
sBCMA.high=1 10      8   18.5    12.5      NA

Create KM for OS & PFS, stratified by sBCMA at day 60 (< vs >= median)

Show code

data1 <- dataset %>%
  filter(!is.na(sBCMA.d60))%>%
  filter(Days.to.death.or.relapse.or.DLC>=60) %>%
  mutate(
    sBCMA.high = ifelse(sBCMA.d60 >= median(sBCMA.d60),1,0)
  )

median(data1$sBCMA.d60)

[1] 5.415

Show code

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC-60)/30, relapse) ~ sBCMA.high, data = data1)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Low","High"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC - 60)/30, 
    relapse) ~ sBCMA.high, data = data1)

              n events median 0.95LCL 0.95UCL
sBCMA.high=0 12     12   20.2   10.63      NA
sBCMA.high=1 12     10   11.1    6.07      NA

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC-60)/30, death) ~ sBCMA.high, data = data1)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Low","High"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC - 60)/30, death) ~ 
    sBCMA.high, data = data1)

              n events median 0.95LCL 0.95UCL
sBCMA.high=0 12      8   33.5    30.1      NA
sBCMA.high=1 12     10   17.1    13.5      NA

Create KM for OS & PFS, stratified by BCMA ABC at screening

Show code

data1 <- dataset %>%
  filter(!is.na(bcma_scrn))%>%
  filter(Days.to.death.or.relapse.or.DLC>=60) %>%
  mutate(
    BCMA.high = ifelse(bcma_scrn >= median(bcma_scrn),1,0)
  )

median(data1$bcma_scrn)

[1] 620

Show code

km_PFS <- survfit(Surv((Days.to.death.or.relapse.or.DLC)/30, relapse) ~ BCMA.high, data = data1)
med_PFS <- surv_median(km_PFS)

PFS <- ggsurvplot(km_PFS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Progression-free survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Low","High"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

PFS

Show code

km_PFS

Call: survfit(formula = Surv((Days.to.death.or.relapse.or.DLC)/30, 
    relapse) ~ BCMA.high, data = data1)

             n events median 0.95LCL 0.95UCL
BCMA.high=0 12     12   13.3    8.07      NA
BCMA.high=1 13     11   21.1    8.37      NA

Show code

km_OS <- survfit(Surv((Days.to.death.or.DLC)/30, death) ~ BCMA.high, data = data1)
med_OS <- surv_median(km_OS)

OS <- ggsurvplot(km_OS,
           pval=TRUE,
           conf.int = TRUE,
           risk.table = TRUE, # Add risk table
           fontsize = 6,
           risk.table.col = "strata", # Change risk table color by groups
           tables.height = 0.25,
           tables.theme = theme_cleantable(),
           #surv.median.line = "hv", # Specify median survival
           #ggtheme = theme_bw(), # Change ggplot2 theme
           #palette = c("#E7B800", "#2E9FDF"),
           xlab="Time (months)", ylab = "Overall survival",
           xlim = c(0,65), break.x.by = c(12),
           ylim = c(0,1), break.y.by = c(0.25),
           legend.labs = c("Low","High"),
           legend = "none",
           surv.scale="percent",
           font.main = c(16, "plain", "black"),
           font.x = c(16, "plain", "black"),
           font.y = c(16, "plain", "black"),
           font.caption = c(16, "plain", "black"),
           font.tickslab = c(16, "plain", "black")
)

OS

Show code

km_OS

Call: survfit(formula = Surv((Days.to.death.or.DLC)/30, death) ~ BCMA.high, 
    data = data1)

             n events median 0.95LCL 0.95UCL
BCMA.high=0 12      8   22.8    18.5      NA
BCMA.high=1 13     11   35.5    17.7      NA

Regression analysis

Cox regression

Show code

theme_gtsummary_compact()
preds <- c("HRCA", "cnLOH", "response.category", "intended_cell_dose", "High.tumor.burden","any_bcma_tx","pentaref","triple_ref","bcma_scrn", "sBCMA.d90", "sBCMA.d60", "sbcma_nadir","sbcma_nadir_timepoint")

data1 <- dataset %>%
  mutate(
    response.category = recode(best_response, "scr" = "sCR/CR", "cr" = "sCR/CR", "vgpr" = "VGPR/PR", "pr"="VGPR/PR"),
    sBCMA.d90 = sBCMA.d90 /10,
    sBCMA.d60 = sBCMA.d60 / 10,
    sbcma_nadir = sbcma_nadir / 10,
    bcma_scrn = log10(bcma_scrn)
  )


uv_tab_OS <- tbl_uvregression(
  data1[c(preds)],
  method = coxph,
  y = Surv(data1$Days.to.death.or.DLC, data1$death),
  exponentiate = TRUE
) %>%
  sort_p()

uv_tab_PFS <- tbl_uvregression(
  data1[c(preds)],
  method = coxph,
  y = Surv(data1$Days.to.death.or.relapse.or.DLC, data1$relapse),
  exponentiate = TRUE
)


tbl_merge(list(uv_tab_OS, uv_tab_PFS), tab_spanner = c("**OS**", "**PFS**"))

Characteristic	OS				PFS
Characteristic	N	HR¹	95% CI¹	p-value	N	HR¹	95% CI¹	p-value
sBCMA.d90	21	3.21	1.39, 7.42	0.006	21	2.59	1.02, 6.55	0.045
sBCMA.d60	24	1.50	1.12, 2.02	0.007	24	1.53	1.07, 2.19	0.019
sbcma_nadir	22	1.42	1.05, 1.93	0.024	22	1.58	1.00, 2.50	0.052
response.category	25				25
sCR/CR		—	—			—	—
VGPR/PR		3.28	1.16, 9.25	0.025		1.95	0.80, 4.74	0.14
High.tumor.burden	25	2.87	0.93, 8.91	0.067	25	2.58	0.92, 7.28	0.073
bcma_scrn	25	0.21	0.03, 1.58	0.13	25	0.35	0.06, 1.96	0.2
intended_cell_dose	25				25
Dose level 1 - 50E6		—	—			—	—
Dose level 2 - 150E6		0.53	0.14, 1.98	0.3		0.34	0.10, 1.17	0.087
Dose level 3 - 300E6		1.04	0.29, 3.74	>0.9		0.45	0.14, 1.45	0.2
Dose level 4 - 450E6		2.96	0.62, 14.1	0.2		2.98	0.66, 13.5	0.2
sbcma_nadir_timepoint	22	1.00	0.99, 1.00	0.4	22	1.00	0.99, 1.00	0.5
cnLOH	22	1.23	0.42, 3.59	0.7	22	0.99	0.36, 2.69	>0.9
pentaref	25	0.90	0.33, 2.45	0.8	25	0.76	0.31, 1.83	0.5
triple_ref	25	0.89	0.29, 2.75	0.8	25	0.68	0.25, 1.85	0.4
any_bcma_tx	25	1.12	0.25, 4.97	0.9	25	1.96	0.56, 6.91	0.3
HRCA	25	0.98	0.37, 2.60	>0.9	25	1.30	0.55, 3.03	0.5
¹ HR = Hazard Ratio, CI = Confidence Interval

Scatterplot

TTR by BCMA ABC at baseline in relapsed patients

Show code

data1 <- dataset %>%
  filter(relapse == 1) %>%
  transmute(
    bcma_scrn = log(bcma_scrn),
    time.to.relapse = Days.to.death.or.relapse.or.DLC/30
  )

ggplot(data1, aes(x=time.to.relapse, y = bcma_scrn)) + 
  geom_point(shape = 16) + # Add points to the plot with shape 16 (i.e., basic circle)
  theme_classic() + # Use the minimal theme
  geom_smooth(aes(x = time.to.relapse, y=bcma_scrn), method = "lm", se = TRUE, color = "#104a8e", inherit.aes = FALSE, linetype = "solid", linewidth = 0.5)+
  stat_cor(method = "pearson", label.x = 3, label.y = 9) +
  labs(x = "Time to relapse (months)", y = "log(BCMA ABC)")

Probability plots

Hazard of OS/PFS vs sBCMA d90

Show code

data1 <- dataset %>%
  select(sBCMA.d90, Days.to.death.or.DLC, death, Days.to.death.or.relapse.or.DLC, relapse)

dd <- datadist(data1)
options(datadist='dd')



fit1 <- cph(Surv(data1$Days.to.death.or.DLC, data1$death) ~ sBCMA.d90, data=data1)
df1 <- data.frame(Predict(fit1, sBCMA.d90, fun=function(x) exp(x) ) )


fit2 <- cph(Surv(data1$Days.to.death.or.relapse.or.DLC, data1$relapse) ~ sBCMA.d90, data=data1)
df2 <- data.frame(Predict(fit2, sBCMA.d90, fun=function(x) exp(x) ) )

ggplot() + 
  geom_line(data=df1, aes(x=sBCMA.d90,y=yhat, color = "OS")) + geom_ribbon( data=df1, aes(x=sBCMA.d90,ymin=lower, ymax=upper),alpha=0.10)+
  geom_line(data=df2, aes(x=sBCMA.d90,y=yhat, color ="PFS" )) + geom_ribbon(data = df2, aes(x=sBCMA.d90,ymin=lower, ymax=upper),alpha=0.10)+
  labs(x = "sBCMA d90",
       y = "Hazard")+
  theme_classic() +
  scale_y_continuous(breaks=c(1,2,3,4,5,6,7,8, 9)) + 
  #scale_x_continuous(breaks=c(0.25,0.5,0.75,1.0,1.25,1.5,1.75), limits=c(0.25, 1.75)) + 
  theme(legend.position = c(0, 1),legend.justification = c(-7.5, 1.1))+
  scale_color_manual(values = c ("steelblue","coral2")) + 
  theme(title = element_text(size=20,face="bold"),legend.title=element_blank())

Hazard of OS/PFS vs screening BCMA ABC

Show code

data1 <- dataset %>%
  select(bcma_scrn, Days.to.death.or.DLC, death, Days.to.death.or.relapse.or.DLC, relapse)

dd <- datadist(data1)
options(datadist='dd')



fit1 <- cph(Surv(data1$Days.to.death.or.DLC, data1$death) ~ rcs(bcma_scrn,3), data=data1)
df1 <- data.frame(Predict(fit1, bcma_scrn, fun=function(x) exp(x) ) )


fit2 <- cph(Surv(data1$Days.to.death.or.relapse.or.DLC, data1$relapse) ~ rcs(bcma_scrn,3), data=data1)
df2 <- data.frame(Predict(fit2, bcma_scrn, fun=function(x) exp(x) ) )

ggplot() + 
  geom_line(data=df1, aes(x=bcma_scrn,y=yhat, color = "OS")) + geom_ribbon( data=df1, aes(x=bcma_scrn,ymin=lower, ymax=upper),alpha=0.10)+
  geom_line(data=df2, aes(x=bcma_scrn,y=yhat, color ="PFS" )) + geom_ribbon(data = df2, aes(x=bcma_scrn,ymin=lower, ymax=upper),alpha=0.10)+
  labs(x = "BCMA screen",
       y = "Hazard")+
  theme_bw() +
  #scale_y_continuous(breaks=c(0,0.2,0.4,0.6,0.8,1.0), limits=c(0, 1)) + 
  #scale_x_continuous(breaks=c(0.25,0.5,0.75,1.0,1.25,1.5,1.75), limits=c(0.25, 1.75)) + 
  theme(legend.position = c(0, 1),legend.justification = c(-8, 1.1))+
  scale_color_manual(values = c ("steelblue","coral2")) + 
  theme(title = element_text(size=20,face="bold"),legend.title=element_blank())

Hazard of PFS vs screening BCMA ABC

Show code

data1 <- dataset %>%
  select(bcma_scrn, Days.to.death.or.relapse.or.DLC, relapse)



dd <- datadist(data1)
options(datadist='dd')

S <- Surv(data1$Days.to.death.or.relapse.or.DLC, data1$relapse)
f <- cph(S ~ rcs(bcma_scrn,3), x=TRUE, y=TRUE,surv=TRUE,data=data1)

model <- Predict(f, bcma_scrn, fun=function(x) exp(x) )

ggplot(as.data.frame(model),aes(x=bcma_scrn, y=yhat)) + 
  geom_ribbon(data = model, aes(ymin=lower, ymax=upper), alpha=0.2, linetype=0) + 
  geom_line() + 
  theme_bw() + 
  labs(
    x = "Screening BCMA",
    y = "PFS (hazard)") +
  theme(title = element_text(size=16,face="bold"),legend.title=element_blank())

Hazard of PFS vs sBCMA nadir

Show code

data1 <- dataset %>%
  select(sbcma_nadir, Days.to.death.or.relapse.or.DLC, relapse) %>%
  filter(!is.na(sbcma_nadir))



dd <- datadist(data1)
options(datadist='dd')

S <- Surv(data1$Days.to.death.or.relapse.or.DLC, data1$relapse)
f <- cph(S ~ sbcma_nadir, x=TRUE, y=TRUE,surv=TRUE,data=data1)

model <- Predict(f, sbcma_nadir, fun=function(x) exp(x) )

ggplot(as.data.frame(model),aes(x=sbcma_nadir, y=yhat)) + 
  geom_ribbon(data = model, aes(ymin=lower, ymax=upper), alpha=0.2, linetype=0) + 
  geom_line() + 
  theme_bw() + 
  labs(
    x = "sBCMA nadir",
    y = "PFS (hazard)") +
  theme(title = element_text(size=16,face="bold"),legend.title=element_blank())

Hazard of PFS vs time to sBCMA nadir

Show code

data1 <- dataset %>%
  select(sbcma_nadir_timepoint, Days.to.death.or.relapse.or.DLC, relapse) %>%
  filter(!is.na(sbcma_nadir_timepoint))



dd <- datadist(data1)
options(datadist='dd')

S <- Surv(data1$Days.to.death.or.relapse.or.DLC, data1$relapse)
f <- cph(S ~ sbcma_nadir_timepoint, x=TRUE, y=TRUE,surv=TRUE,data=data1)

model <- Predict(f, sbcma_nadir_timepoint, fun=function(x) exp(x) )

ggplot(as.data.frame(model),aes(x=sbcma_nadir_timepoint, y=yhat)) + 
  geom_ribbon(data = model, aes(ymin=lower, ymax=upper), alpha=0.2, linetype=0) + 
  geom_line() + 
  theme_bw() + 
  labs(
    x = "Time to sBCMA nadir (days)",
    y = "PFS (hazard)") +
  theme(title = element_text(size=16,face="bold"),legend.title=element_blank())

9762 analysis

Setup and data wrangling

Load necessary packages

Read in the data

Data tables

Table 1 reproduction

Swimmer’s plot

Figure 1: Swimmer’s plot

Supplemental figure

sBCMA plots

Survival analysis

Length of follow-up

Create KM for OS and PFS

PFS and OS on the same plot

Create KM for DOR

Stratified survival analysis

Create KM for OS and PFS, stratify by HRCA

KM of OS & PFS, stratified by cell dose

KM of PFS, dose level 1-3 vs 4

Create KM for OS, stratified by sCR/CR vs VGPR/PR

Create KM for OS & PFS, stratified by treatment lines <8 vs >=8

Create KM for OS & PFS, stratified by hx of prior allo

Create KM for OS & PFS, stratified by receipt of prior BCMA-directed therapy

Create KM for OS & PFS, stratified by triple refractory disease

Create KM for OS & PFS, stratified by penta refractory disease

Create KM for OS & PFS, stratified by tumor burden (BMPCs <=30 vs >30)

Create KM for OS & PFS, stratified by sBCMA at day 90 (< vs >= median)

Create KM for OS & PFS, stratified by sBCMA at day 60 (< vs >= median)

Create KM for OS & PFS, stratified by BCMA ABC at screening

Regression analysis

Cox regression

Scatterplot

TTR by BCMA ABC at baseline in relapsed patients

Probability plots

Hazard of OS/PFS vs sBCMA d90

Hazard of OS/PFS vs screening BCMA ABC

Hazard of PFS vs screening BCMA ABC

Hazard of PFS vs sBCMA nadir

Hazard of PFS vs time to sBCMA nadir